Introduction
To examine relationship between sexual behaviors and practices and HIV prevalence over a period of time among males and females ages 15 to 24 years in
Zimbabwe, I conducted an in-depth secondary data analysis of two ZDHSs for 2005/2006 and 2010/2011. In this study, I used a quantitative methods paradigm and I studied HIV serostatus as the outcome variable. I conducted bivariate analysis using chi-square to examine the association between the IVs and DVs and a logistic regression analysis to predict effects of the IVs on the DVs.
In 2005/06, the odds of being HIV positive was 2.54 times higher among females ages 15 to 19 years and was 1.02 times higher among females ages 20 to 24 years with increasing number of total lifetime sexual partners. The odds of being HIV positive was 0.46 times higher among males ages 20 to 24 years who did not use condoms during last sex compared with those who used condoms. In addition, the odds of being HIV positive was 3.80 times higher among males ages 20 to 24 years and 1.57 times higher among females ages 20 to 24 years formerly in union (divorced, widowed, separated), whereas the odds of being HIV positive among females ages 15 to 19 years decreased by 0.15 for those in union and by 0.45 for those previously in union.
In 2010/11, the odds of being HIV positive among females ages 20 to 24 years was 1.60 times higher with increased total number of lifetime sexual partners compared with those who did not have sex and was 0.75 times more among females ages 20 to 24 years who had first sex by age of 14 years compared with those who never had sex. The
odds of being HIV positive increased by 2.54 among males ages 20 to 24 years in union; increased by 1.79 among females ages 15 to 19 years, and by 1.66 among females ages 20 to 24 years staying in urban areas. The odds of being HIV positive decreased by 0.54 among females ages 20 to 24 years with higher education status. Socioeconomic status did not affect the odds of being HIV positive for all males and females in both years, although a statistically significant association between socioeconomic status and HIV serostatus was observed among females ages 20 to 24 years in 2005/06.
Interpretation of Findings
The findings from this study are in line with the findings from most of the studies conducted worldwide. However, the results in this study are more precise and portray a change in trends of significant predictors of the outcome variables and have further disaggregated the findings among those ages 15 to 19 years (adolescents) and those 20 to 24 years. I present the interpretation of findings from this study and a comparison with existing literature in accordance to the three research questions.
Research Question 1
RQ1 asks the following question: What is the association between sexual behaviors and practices and HIV serostatus in males and females ages 15 to 24 years at two different points in time in Zimbabwe? The sexual behaviors included age at sexual debut, number of lifetime sexual partners, condom use at last sex, and transactional sex/paid sex.
In 2005/06, HIV serostatus was statistically significantly predicted by total
number of lifetime sexual partners among females ages 15 to 19 years and 20 to 24 years, whereas for males, HIV serostatus was predicted by condom use during last sex among
those ages 20 to 24 years. Sexual practices were not a statistically significant predictor of HIV serostatus among males ages 15 to 19 years. The odds of being HIV positive was 2.54 times higher in females ages 15 to 19 years who had more than one lifetime sexual partner (OR = 2.54, 95% CI = [1.70, 3.78], p < .001), whereas for females ages 20 to 24 years, the odds of being HIV positive was 1.02 times greater (OR = 1.02, 95% CI = [1.00, 1.04], p < .05) in 2005/06. The odds of being HIV positive for males ages 20 to 24 years reduced by 0.46 times in those who used condoms during the last sexual encounter (OR = .464, 95% CI = [.253, .854], p < .05) compared with those that did not use condoms in 2005/06.
This is in line with a secondary analysis conducted in Uganda among young people ages 15 to 24 years which showed that having more than one sexual partner significantly increased the risk of HIV infection [OR = 1.94; 95% CI (1.42-2.65), p = <0.01] while condom use [OR = 0.54; (95% CI: 0.41-0.69)] provided a protective effect against HIV (Chimoyi & Musege, 2014). Kembo (2013) in a multivariate analysis conducted in Zambia also found a statistically significant elevated odds of HIV infection of 1.568 times among young persons ages 15 to 24 years with two or more sex partners in the past 12 months preceding the survey compared to their counterparts with no sex partners in the same period of time. Similarly, Bello et al (2011) found that the proximate factors most consistent with underpinning HIV risk reduction in young people of Malawi included reductions in the proportion of men that had several sexual partners and an increase in condom use. This is consistent with findings from a study in Uganda where having three to five lifetime sexual partners was associated with increased odds (OR
adjusted = 2.12, 95% CI 1.50-3.03) of testing positive for HIV among young women (Choudhry et al., 2015).
On the other hand, in 2010/11, sexual practices was a predictor of HIV serostatus only for females 20 to 24 years. The total number of lifetime sexual partners remained a statistically significant predictor of HIV among females 20 to 24 years only and ceased to be a predictor for females 15 to 19 years old. A new statistically significant predictor of HIV serostatus of age at first sex was also observed among females ages 20 to 24 years which was not the case in 2005/06. The odds of being HIV positive was 1.60 times greater (OR = 1.60, 95% CI = [1.31, 1.96], p < .001) in females ages 20 to 24 years who had more than one lifetime sexual partner and was .75 times greater in females 20 to 24 years that had first sex below 14 years old (OR = .753, 95% CI = [.596, .952]) compared to those who never had sex. This is in accordance with a secondary analysis of a cross sectional study conducted in Uganda where a young coital debut (below 14 years) increased the risk of HIV among young people by 58% and this risk was seen to decrease as the age at sex debut increased (Chimoyi & Musege, 2014).
This observation is also consistent with findings from a secondary analysis study by Hallet, Lewis and Lopman (2007) in their study in rural Zimbabwe. They observed that females who started sex at earlier ages were more prone to HIV infection than their peers who had their sexual debut at later ages. A multivariate analysis conducted in Zambia by Kembo (2013) also showed that young persons whose age at first sexual intercourse was less than 14 years were significantly associated with 2.696 times more risk of HIV infection relative to their peers whose age at first sexual intercourse was 20 to 24 years (p = 0.000).
The increased risk to HIV infection among females who engage in early sexual relationships may likely be due to having more sexual partners than their counterparts whose sexual debut occurs at an older age.
On the contrary, a study conducted in Malawi by Bello et al. (2011) indicated that changes in the timing of starting sex was less likely to play a substantial role in reducing HIV risk overall among young people. Although transactional sex was not a significant predictor of HIV serostatus among young people in Zimbabwe in my study, Bello et al. (2011) indicated that transactional sex was positively associated with increased risk of HIV infection [OR = 4.14; 95%CI (1.31-13.13), p = 0.02] among young people in Malawi.
The consistent trend in predictors of HIV infection among females ages 20 to 24 years in 2005/06 and 2010/11 may suggest that no behavioral change was observed regardless of HIV interventions that would have been implemented between the two surveys to influence their reduction in number of sexual partners and yet a change was observed among females ages 15 to 19 years where having more than one lifetime partner ceased to be a significant predictor of HIV infection among them in 2010/11. This
indicates that there may be a need for interventions that will empower females 20 to 24 years to reduce their number of sexual partners and increase condom use if HIV
prevalence is to be reduced among females ages 20 to 24 years. The proximate factors underpinning reduction in HIV incidence among young people in Malawi was reduction in the proportion of young men with several sexual partners and an increase in condom use (Bello et al., 2011). In addition, the fact that age at first sex became a significant predictor of HIV status among females in 2010/11 may suggest the need to have
programs that target the females earlier than the age of 14 years. HIV prevention
programs targeted at young persons ages 15 to 24 years should provide invigorated focus on age at sexual debut, number of sexual partners, and condom use so as to mitigate these predisposing factors for HIV infection (Kembo, 2012).
Furthermore, the results showed an upward trend in HIV positive serostatus among all males ages 15 to 19 years from 3.1% in 2005/06 to 3.4% in 2010/11 and yet sexual practices was not a statistically significant predictor of HIV serostatus among this group in both survey years. An increase in HIV prevalence was noted among those 15 to 19 years old males who never had sex from 2.9% in 2005/06 to 3.7% in 2010/11 while a decrease in HIV prevalence was noted among 15 to 19 year males that had sex by age 16 from 4.8% in 2005/06 to 0.9% in 2010/11. The fact that sexual practices was not a
statistically significant predictor of HIV status among young males 15 to 19 years in both 2005/06 and 2010/11 suggests the possibility that most of the males in that age group who were HIV positive might have acquired HIV from their mothers at birth and grew to adolescence while HIV positive. This could be linked to the increase in number of children living with HIV initiated on antiretroviral treatment over the past 10 years and have grown to adolescence in Zimbabwe. About half of children born with HIV will die before the age of 2 years and about half will die at a median age of 16 years if they are not given ART (Williams, 2011). These results suggest the need to focus on interventions that will reduce the risk of HIV transmission from mother to child to ensure that children are born HIV free which will eventually contribute to reduced HIV prevalence as the children grow to adolescence. Therefore the results suggest that interventions for
adolescent boys and adolescent girls may not be the same since their predictors of HIV serostatus are different.
There may also be need to have a further in-depth study on adolescents (15 to 19 years) living with HIV to determine the extent to which their HIV serostatus is linked to their mothers. These results therefore suggest that programs and interventions to reduce risk to HIV infections should be age and gender disaggregated taking into consideration heterogeneity of young people (Hargreaves et al., 2015). These results further emphasize the importance of delaying sexual debut, hence suggesting that programs should
empower young people, especially females, to delay sex or consistently have protected sex. Implementing behavioral interventions to delay sexual debut and improve condom use can help to reduce the transmission of HIV among youth (Scott-Sheldon et al., 2013). Efforts to control the spread of HIV/AIDS among young people should therefore focus on eliminating sexual practices that have shown to propagate HIV risk.
Research Question 2
RQ2 asks the following question: What is the association between socio- demographic determinants and HIV serostatus in males and females ages 15 to 24 years at two different points in time in Zimbabwe?
In 2005/06, HIV positive serostatus was statistically significantly predicted by being formerly in a union (divorced, widowed, separated) among males and females ages 20 to 24 years and by being in union and previously in a union among females ages 15 to 19 years. The odds of being HIV positive was 3.80 times greater in males ages 20 to 24 years (OR = 3.80, 95% CI = [2.61, 5.52], p<.001) and 1.57 times greater among females
ages 20 to 24 years (OR = 1.57, 95% CI = [1.25, 1.98], p<.001) who were formerly in a union (divorced, widowed, separated) while the odds of being HIV positive decreased by .15 times among females ages 15 to 19 years who were in union (OR = .15, 95% CI = [.07, .30], p<.001) and was .45 times less among females ages 15 to 19 years who were previously in a union (OR = .45, 95% CI = [.22, .92], p<.05) . However, in 2010/11 marital status was a statistically significant predictor of HIV positive status only in males ages 20 to 24 years and ceased to be a statistically significant predictor for females of same age group. The odds of being HIV positive was 2.54 times greater among males ages 20 to 24 years who were in a union (OR = 2.54, 95% CI = [1.60, 4.05]). On the other hand, marital status was not a statistically significant predictor of HIV positive status among males ages 15 to 19 years in both surveys.
The results for females and males ages 20 to 24 years are in line with findings from a study conducted by Chimoyi & Musege (2014) in Uganda, where young people currently [AOR =3.64; (95%CI; 1.25-10.27)] and previously married [AOR = 5.62; (95%CI: 1.52- 20.75)] had significantly higher likelihood of HIV infections relative to those who were never married but contradict these findings for females ages 15 to 19 years. Kembo (2013) in a study conducted in Zambia also reported that young persons ages 15 to 24 years who were divorced, widowed or not living together had significantly increased risk compared to their never-married counterparts, and that marital status was significantly associated with HIV serostatus for the young persons.
Similar findings were reported by Clark, Bruce and Dude (2006) in a study in Africa and Latin America that married young persons ages 15 to 24 years had a higher risk of HIV
infection as compared with their sexually active unmarried peers. The increased likelihood of being HIV infected for males and females ages 20 to 24 years previously in marriage (divorced, widowed, separated) or currently in marriage could be that they most probably had entered marriage at younger ages and unprotected sex increased their risks and in some instances their spouses might have died of HIV or being in previously married exposed them to having more sexual partners than single individuals. However, it should be noted that the results from the reported studies generalized the risk associated with marital status to all 15 to 24 years yet this study has shown that these results are different once further age disaggregation separating adolescents is done.
The observed consistent trends in this study on HIV serostatus predictors among females and males 20 to 24 years in previous union in 2005/06 and those in union in 2010/11 suggests the need for interventions that targets couples to reduce risk to HIV transmission in marriage. This study further suggest that programs and interventions to reduce HIV incidence among young people should focus on widowed and divorced young people as a unique group. Young persons ages 15 to 24 years who are divorced, separated and widowed have significantly higher risks of HIV infection than their unmarried peers therefore programs and interventions for the control of HIV and AIDS should also focus on widowed and divorced young persons ages 15 to 24 years and promote appropriate prevention strategies such as condom use (Kembo, 2012).
Furthermore, the fact that the reported studies lumped young people ages 15 to 24 years suggests that the situation is the same for all of them and yet this study has shown that marital status was not a significant predictor on being HIV positive serostatus for
males and females ages 15 to 19 years. This might be due to the fact that they may not have stayed in marriage long time and their exposure to HIV is reduced. These results emphasize the need to separate adolescents from young adults (20 to 24 years) and disaggregating females and males when analyzing data and developing interventions since their predictors of HIV differ. With the demographic shifts in the adolescent population (expected to double in 2050) in sub-Saharan Africa, a region where HIV infections are also highest, and adolescents already account for 23% of the current population, it is vital that the post-2015 agenda takes into account age-appropriate HIV related interventions aimed at reducing risk, vulnerability, morbidity, and mortality among them (Idele et al., 2014).
In 2010/11, a new statistically significant sociodemographic predictors (place of residence and education status) of HIV serostatus emerged among females which was not a predictor in 2005/06. The odds of being HIV positive was 1.79 times greater in females ages 15 to 19 years (OR = 1.79, 95% CI = [1.07, 2.97]) and 1.66 times greater among females 20 to 24 years (OR = 1.66, 95% CI = [1.17, 2.35]) staying in urban areas compared to their rural counterparts. These results are consistent with findings from a comparative study by Gonese et al. (2010) in Zimbabwe which showed that HIV prevalence was higher in the urban areas (17.8%, 16.5%–19.1%) and lowest in the rural areas (15.1%, CI 17.0%– 18.8%). Although Bello et al, (2011) reported that HIV prevalence dropped from 26% to 15% in urban areas among pregnant women and 40% among those ages 15 to 24 years, compared to no decline among those in rural areas, the HIV prevalence still remained higher among females in urban than those in rural areas (12%). Similarly, Sandoy et al.
(2007) reported that the risk of acquiring HIV infection in 2003 compared to 1995 (AOR) among sexually active young people was higher at 0.35 times [0.28–0.45] for urban females, compared to 0.27 times [95%CI 0.11–0.68] for rural females. Bärnighausen et al. (2007) also reported that urban residence was associated with a 65% increase in the hazard of HIV sero-conversion (p = 0.012) among young people in South Africa. There is therefore need to determine risk factors that increase the HIV vulnerability among females in urban compared to their rural counterparts.
In this secondary analysis, education status also became a statistically significant predictor of HIV serostatus among females ages 20 to 24 years in 2010/11. The odds of being HIV positive was .54 times lower among females ages 20 to 24 years with high education status among (OR = .54, 95% CI = [.39,.77] compared to those with no education. These results are consistent with findings from a multivariable survival analysis conducted by Bärnighausen et al. (2007) in South Africa, who reported that one additional year of education reduced the hazard of HIV seroconversion by 7%. (p = 0.017). Similar findings were reported by Msuya et al. (2007) in study conducted in Tanzania that decline in the HIV prevalence was significant among those who had completed primary education (p < 0.005). Chimoyi & Musege (2014) also reported that higher education (tertiary) was protective against HIV infection by decreasing the likelihood significantly by 68% [OR = 0.34; 95% CI (0.17-0.68), p = 0.02] among young people in Uganda. However, Hargreaves et al. (2002) in their secondary analysis of data on socio-demographic factors and HIV infection from 14 nationally representative surveys of young people ages 15 to 24 (seven countries, two surveys each, 4 to 8 years apart) reported varied results from different
countries. HIV prevalence among young women was lower in the secondary educated groups than primary educated groups in Lesotho, Kenya and Zimbabwe but HIV prevalence was higher in more educated groups in Malawi and Ethiopia and was lowest among those with primary education in Rwanda (Hargreaves et al., 2002).
The results from this study suggest that education could be a protective factor to HIV infection among young females. This could be attributed to the information received in school that empowers them to abstain or practice safer sex or the fact that they are busy with school, they have less free time to have sex. This therefore suggests that programs that promote retention of females in school should be developed or enhanced to reduce the risk to HIV among females. Results from a study by Bärnighausen et al. (2007) showed that increasing educational attainment in the general population may lower HIV incidence. However, the results of this study should be carefully concluded since the lower likelihood levels observed in this study may suggest that there is no strong evidence that education is a predictor of HIV positive serostatus among females in Zimbabwe. This is also in line